Studying B cell secretion at single-cell resolution: Polyfunctionality and Virus-specific antibody discovery

Human and murine B cells are a central component of the adaptive immune system, primarily recognized for their ability to secrete pathogen-specific antibodies during infections or immunizations. However, their functions extend beyond antibody production, encompassing functions of innate immunity, like the secretion of cytokines. Cytokines are crucial signaling molecules essential for regulating and activating immunity, but they are also involved in several inflammatory and auto-immune-like diseases. B cells can induce these disease states, if they become dysregulated, by releasing specific cytokines and large amounts of antibodies. Such release activates other cell types and may cause a chain reaction of inflammation and tissue damage. More recently, B cells have been suggested to be even polyfunctional, capable of secreting both groups of proteins simultaneously. Therefore, it is crucial to understand how both cytokine and antibody secretion function in B cells and how these processes are regulated. However, the existence and functional heterogeneity of these cells, including their secretion patterns, dynamics, and specificity, especially at the single cell level, are not well characterized. To address these gaps, we developed a technology to analyze polyfunctionality in human activated B cells and applied it to study secretion patterns under various conditions. In this thesis we outline the immunological context that leads to the generation of cytokine and antibody expression by B cells. In particular, we introduce advanced techniques and developments of micro devices, specifically designed for single-cell analysis of these cells. We present the development of a technology to measure secretion from single B cells in 2D microfluidic arrays by tracking droplets and cells over-time enabling dynamic secretion analysis. By leveraging techniques of contrastive learning and optimal transport, we achieved highly accurate tracking for cells with both moderate and large movement. Subsequently, we introduce our technology for measuring polyfunctionality at the single-cell level, focusing first on a proof-of-principle for the secretion of IL-1β, IL-6, and TNF-α by human immune cells in general. Finally, we present a platform to assess the frequency, concentration, dynamic patterns, and polyfunctionality of human activated B cells secreting cytokines and antibodies. Our findings indicate that B cells exhibit distinct polyfunctional subsets according to the type of stimulus, its duration, and the specific proteins secreted. In particular, high or rapid secretion of IL-6 and GM-CSF, key pro-inflammatory cytokines, was associated with polyfunctionality and CD19 expression. CD19 was also found to differentially influence the secretion of cytokines and antibodies, suggesting its role as an activation regulator. Interestingly, while cytokine secretion was accelerated with increased CD19 expression, antibody secretion was delayed, correlated with the duration of stimulation. In the second part of this work, we investigated the ability of antibodies to bind pathogens, an essential aspect of antibody quality. In the context of SARS-CoV-2, antibodies can neutralize the virus by blocking entry, inhibiting fusion or affecting endocytosis. However, SARS-CoV-2 variants have evolved mutations in the spike glycoprotein that promote immune evasion by evading antibody recognition. This not only challenges anti-viral immunity, but also makes their analysis more difficult. To decipher how these mutations affect antibody binding, we employed mutational scanning combined with single-cell analysis to detect antibodies secreted in murine and human B cells. By combining our variant analysis with single droplet sequencing, we could map the reactivity of antibodies against the SARS- CoV-2 receptor-binding domain (RBD) under different conditions and throughout the cellular repertoire. Our analysis, performed on IgG secreted by vaccinated mice, revealed significant reactivity at the monoclonal level, specifically against alpha, delta, and Omicron B.1 variants of SARS-CoV-2. Finally, we also evaluated our approach for its use in the discovery of antibodies from secreting B cells. We identified and isolated potentially cross-reactive SARS- CoV-2 specific B cells from previously infected patients based on the secreted antibody and performed single-cell RNA sequencing to obtain the antibody sequence. This work serves as a proof-of-principle for a matched antigen-antibody analysis. Our findings demonstrate the feasibility of variant-wide analysis directly from secreting plasma cells and memory B cells - an otherwise inaccessible repertoire. This work highlights the potential of single cell analysis to uncover cross-reactive antibodies, identify novel epitopes, and thus improve our understanding of B cell-mediated immunity. Together, the technologies presented in this work hold promise as a research tool to study immunization and infection at high resolution. They allow investigating inflammatory dysregulation of B cells and their polyfunctionality and enable functional antibody discovery for the development of next-generation, antiviral therapeutics.